Scalable and Dynamic Big Data Processing and Service Provision in Edge Cloud Environments

Owing to the exponential growth of connected devices and the large amounts of data produced by such devices, clouds are becoming a bottleneck and cause latency while collecting and processing data and providing associated services [1]. The concept of edge computing has been suggested to solve this scalability problem by moving data centers and computing resources close to the data sources [2]. Locally deployed data centers and computing resources form an edge cloud or a fog that can collect and process big data in a distributed and scalable manner [3]. Recently, low-latency and reliable communication technologies such as 5G have enabled more effective realization of edge cloud environments [4].
Edge clouds are especially useful for efficient, reliable, and secure data collection and processing for smart cities and smart factories [5]. Figure 1 shows an overview of big data-driven service provision in a smart city environment. In this environment, edge devices such as cars and sensors on roads collect data and accumulate them in nearby edge clouds. Then, the services running on an edge cloud process them locally and provide localized service capabilities to nearby users. Such services can be dynamically deployed to edge clouds according to user needs and various environmental contexts.